Article
Computer Science, Artificial Intelligence
Deshuai Zheng, Jin Yan, Tao Xue, Yong Liu
Summary: Task planning is essential for robot multi-task manipulations. Language-based methods provide practicality in receiving commands from humans and require low-cost labeled data. To overcome the limitations of existing methods, we propose a knowledge-based approach called Recurrent Graph Convolutional Network (RGCN) that leverages knowledge graph data and historical predictions. Our approach achieves a task planning success rate of 95.7%, surpassing the best baseline method significantly.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Zhongli Wang, Guohui Tian
Summary: This article proposes a task-oriented robot cognitive manipulation planning method using affordance segmentation and logic reasoning, which can provide robots with semantic reasoning skills to manipulate appropriate parts of an object according to different tasks. The method utilizes a convolutional neural network based on the attention mechanism to obtain object affordance and constructs object/task ontologies for the management of objects and tasks. By establishing object-task affordances through causal probability logic, the method can reason manipulation regions' configuration for the intended task with the help of the Dempster-Shafer theory. Experimental results demonstrate that this method effectively improves the cognitive manipulation ability of robots and enhances their performance in various tasks.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Physics, Multidisciplinary
Priyam Parashar, Ashok K. K. Goel, Henrik I. I. Christensen
Summary: This paper explores the complex problem of robots assembling objects and discusses how to improve system performance by combining low-level information and high-level expectations. By using meta reasoning architecture and perceptual expectations, a dual encoding approach is proposed to determine nominal scenarios during task progress. Results show that in practice, considering both low-level information and high-level expectations performs better than using them separately.
FRONTIERS IN PHYSICS
(2022)
Article
Robotics
Paolo Forte, Anna Mannucci, Henrik Andreasson, Federico Pecora
Summary: The study introduces a framework for handling task assignment, motion planning, coordination, and control of heterogeneous fleets of robots for non-cooperative tasks. It addresses the real-world requirement of asynchronous task posting and offers a safe and efficient solution.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Computer Science, Artificial Intelligence
Hejia Zhang, Shao-Hung Chan, Jie Zhong, Jiaoyang Li, Peter Kolapo, Sven Koenig, Zach Agioutantis, Steven Schafrik, Stefanos Nikolaidis
Summary: In this paper, we propose a solution for multi-robot geometric task-and-motion planning problems. Our approach collects occlusion and reachability information for each robot and builds a graph structure to guide the search for highly effective collaborative task-and-motion plans. Experimental results show that our approach outperforms other methods in terms of planning time, plan length, and number of objects moved, and can be applied to underground mining operations.
Article
Computer Science, Theory & Methods
Luke Antonyshyn, Jefferson Silveira, Sidney Givigi, Joshua Marshall
Summary: With recent advances in mobile robotics, autonomous systems, and artificial intelligence, there is a growing expectation for robots to solve complex problems, particularly in multi-robot systems. Many recent works in the field of combined task and motion planning for multiple mobile robots have integrated task and motion planning to address these complex tasks. By categorizing works based on their underlying problem representations, the authors survey the recent contributions and propose a taxonomy for task and motion planning applicable to both multi-robot and single-robot systems.
ACM COMPUTING SURVEYS
(2023)
Article
Automation & Control Systems
Wenrui Zhao, Weidong Chen
Summary: This work introduces a new hierarchical POMDP framework for object manipulation tasks, which improves planning efficiency by extracting a brief abstract POMDP and includes a learning mechanism for unknown probabilities. The framework is demonstrated with an object fetching task and validated empirically through simulations and experiments.
ROBOTICS AND AUTONOMOUS SYSTEMS
(2021)
Article
Robotics
Yuanfan Xu, Zhaoliang Zhang, Jincheng Yu, Yuan Shen, Yu Wang
Summary: This letter presents a framework to co-optimize robot exploration and task planning in unknown environments. A unified structure called subtask is designed to decompose the exploration and planning phases, and a value function and value-based scheduler are developed to select the appropriate subtask each time. The framework is evaluated in a photo-realistic simulator, achieving a 25%-29% increase in task efficiency.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Robotics
F. Cini, T. Banfi, G. Ciuti, L. Craighero, M. Controzzi
Summary: In human-robot collaboration, the timing of when robots provide information can significantly impact task performance, with timely cues leading to reduced errors and increased efficiency in humans.
Article
Agriculture, Multidisciplinary
Tao Li, Feng Xie, Zhuoqun Zhao, Hui Zhao, Xin Guo, Qingchun Feng
Summary: Robot harvesting is urgently needed in the apple industry due to a decline in agricultural labor. The use of multiple robotic arms in harvesting robots has gained attention for improving efficiency. However, the efficiency and accuracy of fruit positioning hinder the widespread application of multi-arm harvesting robots in orchard production. This paper proposes a multi-arm apple harvesting robot system that focuses on precise perception and multi-arm collaborative control. It introduces a hardware and software integration method, a fruit recognition and localization algorithm, and a task planning method to optimize collaboration efficiency. Field experiments in orchards validate the effectiveness of the robot and its methods.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Robotics
Alessandro Palleschi, George Jose Pollayil, Mathew Jose Pollayil, Manolo Garabini, Lucia Pallottino
Summary: Multi-robot systems are gaining popularity in warehouses and factories due to their ability to provide more efficient and complex task solutions. This article introduces a method based on a hierarchical planning framework to address the complexity in multi-robot systems, and validates the effectiveness of the method through experiments.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2022)
Article
Computer Science, Information Systems
Faiza Gul, Imran Mir, Wan Rahiman, Tauqeer Ul Islam
Summary: This paper proposes a novel framework that integrates deterministic Coordinated Multi-Robot Exploration (CME) and metaheuristic frequency modified Whale Optimization Algorithm (WOA) techniques to perform search exploration imitating the predatory behavior of whales. The framework dynamically adjusts frequency using a statistical objective function to tune exploitation and exploration operators to improve the overall solution optimization.
Article
Mathematics
Ali El Romeh, Seyedali Mirjalili, Faiza Gul
Summary: This study proposes a novel hybrid optimization method called Hybrid Vulture-Coordinated Multi-Robot Exploration (HVCME), which combines Coordinated Multi-Robot Exploration (CME) and African Vultures Optimization Algorithm (AVOA) to optimize the construction of a finite map in multi-robot exploration. Experimental results show that HVCME outperforms four other similar methods, demonstrating its effectiveness in optimizing the construction of a finite map in an unknown indoor environment.
Article
Robotics
Wenyu Liang, Fen Fang, Cihan Acar, Wei Qi Toh, Ying Sun, Qianli Xu, Yan Wu
Summary: A new visuo-tactile feedback-based manipulation planning framework is proposed in this work, which uses multisensory feedback and an attention-guided deep affordance model as perceptual states, and a deep reinforcement learning pipeline. Multiple sensory modalities, including vision and touch, are employed to predict and indicate manipulable regions for objects with similar appearances but different intrinsic properties. The proposed method achieves better accuracy and higher efficiency in object packing task.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Computer Science, Software Engineering
Xuefeng Dai, Jiazhi Wang, Jianqi Zhao, Dahui Li, Zhifeng Yao
Summary: A coordinated strategy based on predicate logic reasoning is proposed in this paper to improve time efficiency for structured environment exploration of multi-robot systems. The strategy assigns the room with maximal utility to the robot with the optimal exploring property and realizes cooperative exploration by considering the room capacity. The coordination takes into account robot performance and environment property and is conducted in a distributed manner. In addition, a triple set named task information is proposed to save computing and storage resources for environment information sharing among robots.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2022)
Article
Computer Science, Information Systems
Gopal Singh Kushwah, Virender Ranga
Summary: This paper proposes a DDoS attack detection system based on an improved Self-adaptive evolutionary extreme learning machine (SaE-ELM), which achieves high detection accuracy on multiple datasets. The system shows significantly improved learning and classification capabilities, outperforming other techniques in performance.
COMPUTERS & SECURITY
(2021)
Article
Telecommunications
Rochak Swami, Mayank Dave, Virender Ranga
Summary: Software-defined networking (SDN) is an advanced technology that provides flexibility and cost-efficiency based on business requirements. This study focuses on the impact of spoofed and non-spoofed TCP-SYN flooding attacks on controller resources in SDN architecture, and proposes a machine learning based intrusion detection system.
WIRELESS PERSONAL COMMUNICATIONS
(2021)
Article
Telecommunications
Ehsan Zadkhosh, Hossein Bahramgiri, Masoud Sabaei
Summary: In this article, an intelligent dynamic routing framework based on SDN architecture with the use of genetic algorithm (GA) for performance optimization is proposed. By extracting CPU, memory, and bandwidth utilization of middleboxes (MBs) as dynamic routing parameters and calculating impact factors (IFs), the network performance is improved.
TRANSACTIONS ON EMERGING TELECOMMUNICATIONS TECHNOLOGIES
(2021)
Article
Telecommunications
Gopal Singh Kushwah, Virender Ranga
Summary: In this paper, a hybrid machine learning model based approach is proposed to detect DDoS attacks in cloud computing. The model utilizes extreme learning machine and adaptive differential evolution to optimize the weights, and analytically determines the connection weights. The performance of the system is evaluated using three state-of-the-art datasets.
WIRELESS PERSONAL COMMUNICATIONS
(2022)
Review
Computer Science, Artificial Intelligence
Ankit Attkan, Virender Ranga
Summary: The Internet of Things (IoT) has gained significant attention in recent years for its ability to improve lifestyles and keep up with technological advancements. IoT edge devices have different technologies and storage formats, requiring secure mutual authentication for data transmission. Blockchain and AI are integrated into IoT networks for enhanced security, with blockchain storing validated session keys and AI providing better adaptability against attacks.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Information Systems
Aditi Zear, Virender Ranga
Summary: This paper presents a solution for UAV-assisted network partition detection and connectivity restoration. The proposed algorithm outperforms existing techniques in terms of recovery time, distance traveled by UAVs, and number of relay nodes deployed. The paper also provides a new framework covering all aspects required for network partition recovery.
Article
Engineering, Electrical & Electronic
Aditi Zear, Virender Ranga, Kriti Bhushan
Summary: This paper introduces an approach for partition detection and recovery in damaged WSANs using unmanned aerial vehicles (UAVs), which outperforms existing algorithms in terms of recovery time, relay nodes, and UAVs' travel distance.
INTERNATIONAL JOURNAL OF COMMUNICATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Ravinder Kumar, Amita Malik, Virender Ranga
Summary: Smart devices connected to the internet, known as the Internet of Things (IoT), have advantages in improving interaction but also raise privacy and security concerns. This study proposes a hybrid Hunger Games Search and Remora Optimization Algorithm (HHGS-ROA) to address security problems in IoT networks, achieving better performance than existing methods.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Engineering, Multidisciplinary
Rochak Swami, Mayank Dave, Virender Ranga
Summary: Software-defined networking (SDN) is a networking paradigm that focuses on decoupling control logic from the data plane, bringing programmability and flexibility to network management. SDN faces security issues such as control plane exhaustion and switch buffer overflow. This paper presents a defense solution implemented in the SDN controller to detect and mitigate spoofed flooding DDoS attacks using statistical measures and existing SDN capabilities.
DEFENCE TECHNOLOGY
(2023)
Article
Computer Science, Information Systems
Aditi Zear, Virender Ranga, Kriti Bhushan
Summary: This paper presents a new solution for coordinated network partition detection and bi-connected inter-partition topology creation in damaged sensor networks using multiple UAVs (UAV-BITS). UAV-BITS employs multiple UAVs to identify network partitions and guide network nodes to self-identify their partitions by creating Connected Component Sets (CCSs) iteratively. During the network recovery process, UAV-BITS creates two-vertex disjoint paths between network partitions to ensure a fault-tolerant network topology. Simulation results show that UAV-BITS provides a better fault-tolerant network topology compared to existing solutions, while also reducing UAV travel distance and detection time.
COMPUTER COMMUNICATIONS
(2023)
Article
Telecommunications
Rochak Swami, Mayank Dave, Virender Ranga
Summary: Software-defined networking (SDN) is a networking paradigm that separates control logic from data plane for increased programmability and flexibility. However, SDN faces security issues such as control plane exhaustion and switch buffer overflow, which can hinder its growth and adoption. In this paper, a defense solution based on moving target defense (MTD) and existing SDN capabilities is proposed to mitigate spoofed flooding DDoS attacks. The solution is implemented in the Ryu controller and evaluated using CPU usage, showing effective mitigation of the attack.
WIRELESS PERSONAL COMMUNICATIONS
(2023)
Review
Telecommunications
Ravinder Kumar, Amita Malik, Virender Ranga
Summary: This paper discusses the security vulnerabilities of the RPL protocol in IoT and proposes solutions. It provides an overview of the IoT stack model and examines the security concerns of the RPL protocol. Different types of attacks and their defense solutions are compared, and the use of emerging technologies like ML, Blockchain, and Cloud Storage for IoT security is explored. The paper also identifies research challenges and suggests future research directions for IoT security.
TRANSACTIONS ON EMERGING TELECOMMUNICATIONS TECHNOLOGIES
(2023)
Article
Computer Science, Information Systems
Ankit Attkan, Virender Ranga, Priyanka Ahlawat
Summary: Over the past decade, IoT has made significant progress in terms of technology exploration, integration, and various applications despite its resource-bound architecture. The challenge lies in running high-end security protocols on Edge devices, which are highly vulnerable to cyber-attacks. To address this, proper mutual authentication and session key establishment are crucial, and a Rubik's cube puzzle-based cryptosystem is proposed in this article for parameter exchange. Incorporating blockchain technology ensures anonymity of token transactions, and a session key pool randomizer is used to prevent probabilistic attacks. The hybrid model generates secure session keys for mutual authentication and reliable data transferring tasks, exhibiting promising results in terms of efficiency, light weightiness, and practical applications.
ACM TRANSACTIONS ON INTERNET OF THINGS
(2023)
Article
Environmental Sciences
Abhishek Verma, Virender Ranga, Dinesh Kumar Vishwakarma
Summary: This research article introduces a new approach for forecasting PM2.5 pollution in Delhi by combining a pre-trained CNN model with a transformer-based model called CATALYST. The proposed CATALYST model incorporates advancements in the transformer-based architecture and utilizes a sliding window training approach to predict future pollution levels. The experiments demonstrate that this model outperforms conventional methods and has the potential to accurately forecast PM2.5 pollution.
ENVIRONMENTAL MONITORING AND ASSESSMENT
(2023)